A Story of Life & Death and Love & War

By

H.M. Hubey

Montclair State University

1Introduction

Looking at the literature on diachronic linguistics, we find a method variously called the “historical method”, or the “comparative method”, and often spoken in conjunction with “Swadesh List”, “lexicostatistics” or “glottochronology”. In short, it is an intuitive method, based on heuristics which could be backed up by probability theory but not explicitly stated or computed. There are some small signs that things are changing Ringe [1992], Manaster-Ramer [1998], Hubey [1994]. The most fundamental idea behind the “comparative method” of historical linguistics is that if we find a set of words in language X which can be changed into a set of words in language Y using “regular sound changes” we can rest assured that it is not due to an accident or chance. From this follows the attempts to decide whether the sound correspondences are “genetic” or “borrowing”. It is this part that causes problems for the linguistics community. The dominant paradigm of the century is that a small subset of words in any language are less prone to be borrowed and hence are evidence for geneticity. These have been more or less formalized in the Swadesh list of 100 or 200 words. Some words which do occur in many languages have been tossed aside as “child/infant language”. Among these are words for family relationships such as ata, atta, ana, anna, papa, apa, api, ati, eke etc. Obviously, it is a great blow to Turkic languages since the Turkic languages have so many of these. For example, the single Turkic language Karachay-Balkar (KB) of the North Caucasus has these words in active use: ata, ana, amma, appa, akka, atta, anna, egeç. Obviously, the historical linguistics field now resembles that of anthropology/paleontology in that it consists basically of lumpers vs splitters. The lumpers are variously called proto-worlders, or nostraticists, the most famous ones being Diakonoff, Dolgopolsky, Shevoroskin, Bomhard, Greenberg, and Ruhlen. The splitters, Indo-Europeanists featured heavily among them, basically cling to the notion that no family/genetic relationships among languages can be discerned after some period of years, usually said to be about 6,000 years although the age of English has recently been changed to 8,000 years [Lass,1997]. No convincing reason for any of these dates has really given anywhere. It is often found that those who deride lexicostatistics are more than happy to use its results (i.e. the 6,000 year age of IE).

In addition, there are other serious problems with this method. Among these are

  1. Some languages do not have a written record stretching back to thousands of years BC so that other methods (even more rigorous ones) must be used.
  2. The first place/language we see a word written in some form is the language that we assume the word belongs to. This obviously works immensely to the advantage of the Afro-Asiatic (Egyptian, Akkadian, Hebrew, etc), Indo-European (Sanskrit, Hittite, Greek, Latin), and Chinese.
  3. The attempts to create 3 kinds of relationships such as genetic, areal, typological suffer from a malady Goethe identified centuries ago: “When an idea is lacking a word can always be found to take its place”. It is as if one person claimed that thunder and lightning occurred together often so there must be a link, and another answered “Hold on there, lightning is electromagnetic radiation and thunder is an acoustic wave. You are mixing up apples and oranges!” Or perhaps, to explicitly paraphrase Goethe’s diagnosis, as if someone said “Well, we know that lightning and thunder occur together. They are called physics universals, and they don’t mean anything more!”
  4. The heuristic rule of regular sound change itself can be self-contradictory or meaningless. There are languages with as few as 13 phonemes (Hawaiian). The languages in the main theater of history have more. And the languages which were isolated for a very long time sometimes have lots more and also possess certain phonetic peculiarities. Therefore if we are projecting backwards to 10,000 years or more we probably should consider languages not having many phonemes and not too many words. In this case, there’s a mathematical concept called the Pigeonhole Principle which should apply. If we have n pigeonholes and n+1 pigeons, then at least one pigeonhole has at least 2 pigeons. Similarly if we are looking at two languages which have n consonants and any consonant can change to any other consonant (ignoring no change) we have n(n-1) possible sound changes. If we therefore find n(n-1)+1 cognates we are guaranteed to have at least 2 examples of the sound change. Even if sound changes were at random we would have a 50% chance of having at least one sound change repeated long before n(n-1) is reached. So unless rigorous probabilistic methods are used, many problems in historical linguistics cannot be resolved via heuristic rules. Some linguists who have attempted such analysis are Cowan [1962], according to whom 3 pairs are sufficient to establish geneticity; Bender [1969] who thinks 2-7 cognate pairs are needed and Greenberg [1960] for whom 3-4 cognates are sufficient. The work by Ringe [1992] and its criticism by Manaster-Ramer [1998] on the same problem are noteworthy. More results can be seen in Hubey [1994].

It seems that for languages for which we do not have huge mass of data and which are spread over 5,000 to 10,000 years or more we have to try a different paradigm for historical linguistics than the one that works for IE and AA languages for which we have many thousands of years of data from many branches. The justification is via the exact same probability laws.

2Clustering Paradigm:

2.1.1.1The root ti=upright, vertical=healthy, alive

Here, two such intuitive paradigms are presented with examples which can be used to counteract the weaknesses of the standard historical method. The scientific or mathematical basis of the method will be presented only via analogies since this is not a forum for such works. On page 38 Anttila’s immensely influential book on historical linguistics we find; “...thus the Sumerian word ti ‘arrow’ was written with a picture of an arrow, but the same sign was also used for ti ‘life’”. From this simple description nothing further can be gleaned. First, we note that there are almost 200 cognates between Sumerian and Turkic, all listed in order of “regular sound correspondance” as required by the heuristics of the present state of linguistic science [Tuna,1990]. Therefore there is no need to imagine that it is an accident, although it can be. One way to show that this is not such an accident is to create an etymology for these words by finding all words related to this word both phonologically and semantically in all the presumably related languages. Since we cannot get anymore information from Sumerian, we have to turn to Ural-Altaic languages. Since Tuna’s words are from Turkic we can look there first. We find, that the word for “alive” in various Turkic languages is: Turkish/Turkmen /diri/, KB /tiri/, Kazakh /türüü/, Uyghur /tirik/, Uzbek /tirik/ [Topçuoglu,1996:5]. In Chuvash it is /çĕrĕ/ [GronbechV,1979:32].

The earliest such word is from Orkhon in the form of /tirg/ (living), also in Koibal /tirig/ [GronbechV,1979:31]. In the same place /türü/ is given as ‘living’ but not the language. We also have Taranchi /tirik/, Kazan /tere/, Yakut /tilin/ (to become living), Altai /tirġis/ (to make a living) [GronbechV,1979:42]. Furthermore in Turkish ‘dik’ means “to plant a sapling’. The word for planting seeds is “ek”. The word for ‘dik’ then by analogy is also used to mean erecting such as erecting a house. The Egyptian obelisk in Istanbul is called “dikili taş”. The word “dik” can be used as a verb as in “dikildi” meaning “he/she stood up(right)”. Hence the root /tik/ in Karachay-Balkar, and the corresponding /dik/ in Turkish have come to mean ‘steep, nearly vertical’. This is likely due to a contraction of the form tiriktirktik like toluktolktok (full, satisfied). Other attested words for ‘steep’ are Koibal /tige/, and Chuvash /çige/ [Gronbech,1979]. The word /tire/ means ‘to lean (often as a support)’ in KB. /Dire/ means the same thing in Turkish. Yakut /tirie/ (to support), and Koibal /tire/ [Gronbech,1979:77] are also a part of this paradigm. Their unrounded versions, /taya/, and /daya/ mean the same thing, and most likely have been re-absorbed from yet another Turkic dialect.

We also have /direk/ meaning ‘pole’ in Turkish. /Terek/ means ‘tree’ in KB and is likely related to ‘direk’. It is /däräkh/ in Uygur, and /daraxt/ in Uzbek [Topçuoglu,1996:157]. It was /taru/ in Hittite and obviously ‘tree’ in English. The KB word /tınç/ (meaning healthy, quiet) probably also belongs to this cluster. The Turkish version of the word, as expected is /dinç/. Turkic /tın/ (life) [Tuna,1990:21], and KB /tunçuk/ (to choke) as well as Yakut /tıın/ (to breathe) and /tıŋa/ (lungs) seem to belong to the same cluster having to do with living and breathing. Perhaps /tın/ > /can/ is a possibility and /can/ is not really from Farsi.

However, there does not yet seem to be any reason yet to connect life and arrow. But we can already see that the words have something to do with “being vertical/upright”. After all, we would expect that back in the days when perhaps only a few thousand words might have been in use, either the people were upright/standing or keeled over and sick or dead. The only other time they were not upright would have been when they were sleeping and another word probably sufficed for that. But what is not clear yet is why arrow should be connected with uprightness/verticality. Vertical things like trees are also alive, but poles /direk/ are not. The word ‘ti’ was gleaned from Sumerian of thousands of years ago. They could have acquired these meanings thousands of years before that. It is a testimony to the diligent work of linguists that such a word can even be recovered. However it is relatively straightforward to find an explanation for the arrow≡life connection. Before the metals age, it would have been impossible to work with large trees the way we do now. The easiest way to make an arrow would have been to start with something as close to an arrow as possible. That would have been a sapling. One would have to only tear it off, scrape off the bark, and then heat treat it to make it straight and then attach a stone point. Hence the relationship between verticality and life. The Uighur, and Karachay-Balkar /tüz/, Turkish /düz/, Kazakh /tuwra/, Kyrgyz /tuura/, Tatar /turi/ and Uzbek /to’g’ri/ meaning ‘straight’ (as an arrow?) are likely from the same root. Strangely enough Turkish also has /dogru/ (straight ahead, or true) and KB has /tuwra/ (straight ahead or true). These words seem like they are the l~r equivalent of ş~z Turkic /tüz/düz/. Others likely have both variants from the same root.

2.1.1.2Aw/Awuş

If the semantic changes from /ti/ (upright) to /ti/ (life) is not convincing, or seems farfetched we can look into the attested words of Karachay-Balkar in which similar changes have occurred; /aw/au/ (to fall over, keel over), /awuş/ (to die), /awruw/ (pain, sickness (i.e. Turkish ağrı)), /arı/ (to tire out (via extension from awruw)); /awur/ (heavy (i.e. Turkish /ağır/)) again via analogy from the fact that the people hefting around heavy things all day would get “ağrı”. The next is not so clear but it is likely from the same root; /aqırın/ or /aqırtın/ meaning “slowly”. Obviously, if you lug heavy things you go slowly. This can be compared to Kyrghyz /aqırın/ (slowly). The strange thing is that /awuş/ corresponds to Sumerian /uş/ (which means ‘to die’) and which is one of the regular sound changes between Sumerian and Turkic that Tuna [1990] has found and which in Turkic is /öl/. The strange thing is that both versions now exist in Karachay-Balkar language. And we have corroborating evidence from Tatar in the form of /avır/ (difficult, heavy); /avırtu/ (pain), and /avıru/ (illness). In Chuvash /yıwer/ (heavy) and Yakut /yaraxan/ are cognates [Gronbech,1979:83]. Chuvash /ıwër/ means ‘to grow tired’ [GronbechV,1979:120] and is cognate with Karachay-Balkar /arı/. Furthermore, there is at least one more example of two such words in existence; Balkar /nıgış/ meaning ‘gathering’ is cognate with both Sumerian /nigin/ and Turkish /yığın/ with n>y being one of the regular Sumerian>Turkic sound changes identified by Tuna [1990:9] the Karachay-Balkar primary version being /cıyın/).

2.1.1.3Cylindrical Pointed Objects, and Piercing

To find related meanings we have to turn to other languages; in Hungarian /tü/ (needle), /tüzni/ (to prick), /tüske/ (thorn), /tör/ (stabbing sword). Once again in what looks like the l~r versions are being preserved Turkish /dürt/ (to push, to shove), KB /türt/ with the same meaning. The word for feather is /tük/ in KB and Azeri, and /tüy/ in Turkish. In KB, and Altay /tük/ is also used for body hair, or feathers. In Kazan Tatar the word is /tök/; in Yakut /tüü/; In Osmanli, /tüs/ was ‘down’. [GronbechV,1979:46]. Hungarian /tur/ (to dig into), and /şur/ (i.e. szur, to prick, stab) are there to remind us of the l~r and ş~z alternation in Turkic languages as well as the shift from plosives such as /t/ to sibilants such as /s/ and /ş/ as does the Turkish ikileme delik-deşik. In fact the word for hole in Turkish should be deşik and not delik since the latter clearly should belong to l~r Turkic. KB /tüy/ and Turkish /döv/ meaning ‘to beat up’ are likely derived from a similar word meaning ‘club, stick’. Turkmen and KB /çüy/ (nail) and Turkish /çivi/ likely from the same root. Even Kazakh /şege/ can be from the same root because the words for hammer contain similar roots. There is the smoking gun (for correctness of these) in Turkish /çekiç/ (hammer), and KB /çögüç/ (hammer) and Tatar /çükeç/ [Topcuoglu,1996:69] are all related to /çüg/çüy/çög/ (nail) and to Turkish /çak/ (to hammer) which in turn is cognate with Sumerian /sag/ [Tuna,1990:11]. The other root for nail /mikh/ seems related to /miz/biz/ (awl) and is probably from the root /kız/ (to heat up) from an earlier form /mir/ which shows up in Sumerian [Tuna,1990:8]. There is Turkish slang /çük/ meaning ‘phallus’ (and which is related to words such as Etruscan /sek/ ‘girl’ [Lahovary,1963:206] (sec in Latin), Sumerian /sig/ ‘to strike’ [Lahovary,1963:295], Sumerian /sig/ ‘to procreate’ [Lahovary,1963:252] or ‘to procreate’ [Lahovary,1963:206] and Turkic /sik/ (penis, to copulate). Probably here also belongs Turkic /siy/ (to urinate). What is obvious even now is that the transition t>s had already occurred in Sumerian. In KB /süy/ also means ‘to love’ which is /sev/ in Turkish but /süyel/ means ‘to stand up’.

2.1.1.4Sew, Bind, and Tie

However, /tik/ means ‘to sew’ in KB, Azeri, Turkmen, Uzbek Uygur and Kyrgyz; it is /dik/ in Turkish, and /tig/ in Kazakh [Topçuoğlu,1996:131]. It is obviously from piercing using a sharp instrument like, a nail, or needle. There is a semantic shift in Orkhan, Yakut and Altai in which /tik/ means ‘to stick’. It is /çik/ in Chuvash, and /tek/ in Kazan Tatar.

In this case we should be able to find words in Turkic which are similar both semantically and phonologically. In fact, they exist in plenty. Turkish /döğme/ (tattoo) is likely from /tü/ (needle) and not from modern /döğ/döv/ (to beat up). Once we make the semantic transition into "to bind" (from "to sew") then we also obtain Turkish /düğüm/ (knot), KB /tüyümçek/ (knot). In the other dialects we have Azeri /düyün/, Kazakh /tüyin/, Kyrghyz /tüyün/, Turkmen /düwün/, Uighur /tügün/, and Uzbek /tugun/. Of course Sumerian /tugdu/ and Old Turkic /tügün/ also belong in this category [Tuna,1990:15]. Even Turkish /düğün/ (wedlock, or should we say wed-knot) probably belongs to this category. As for another way of binding things together (i.e. via buttons) we have KB and Kazakh /tüyme/ and Turkish /düğme/ (button), Kyrgyz /tüymö/, Tatar /töymä/, Uygur /tügmä/ and Uzbek /tugma/ none of them likely from ‘to beat up’ but rather ‘to bind’. Further along the same semantic but not phonological territory we find Kyrghyz /tüy/ (to bind), and Altaic /tüü/ with the same meaning. Kazan /töyön/ (knot), Chuvash /tewa/tewe/, and /tüle/ (‘to bind’ as in ‘düğümle’ meaning ‘to button up’) complement the others. We also have Yakut /tüm/ (to bind), and /tümük/ (knot) as witness that the analysis is along the correct lines [GronbechV,1979:137].

2.1.1.5Tooth, tongue, and other pointed objects

Furthermore the word for ‘tooth’ is Kyrghyz/KB/Uzbek /tiş/ Turkish/Azeri/Turkmen /diş/, Kazakh /tis/, Tatar /teş/, and Uighur /çiş/ all point to the same root with Uighur displaying the t>ç sound change that we saw above. As amazing as it seems the word for ‘tongue’ /til//dil/tel/ also all point to the /ti/ root. Whether the word for tongue (a body part) came before the word for ‘standing up’, or ‘arrow’ is best left for historical linguists to ponder. Although the view that some words are resistant to borrowing apparently holds up in Indo-European with the caveat that we have to exclude some words by making special exceptions such as ‘baby talk’ it does not hold up in other languages [Dixon,1997] and furthermore the exceptions are the best candidates for proto-world, or Nostratic. Connected with the word for tongue naturally is speaking and we have /de/ ‘to say’ which again pops up in Sumerian as /ti/ and which in Old Turkic was /ti/ [Tuna,1990:21]. Strangely enough we also have in English ‘tell’ and ‘say’ displaying the t>s transition. To continue, we have Turkish /diken/ (thorn), KB. /çıġanak/ (thorn) are from the same paradigm. The unrounded versions of /tü/ti/ also exist in KB as /tayak/ and Turkish as /dayak/ (stick,bat, club) again from the same /ti/tü/ root. Furthermore there is Turkish /değnek/ (bat,club) and /değ/ (to touch) KB /tiy/ (to touch). Yakut /tii/ (to reach) is from *ti (to touch)=teg [GronbechV,1979:77] which are clearly related to /tüy/döv/ i.e. instead of "touch" we touch with a little bit more power and it becomes "beat". There is also KB /tiyek/ (buttons on the accordion) from /tiy/ (touch). Sumerian /tuku/ ‘seben, Kleid’ and Turkic /toku/ ‘to hit’ and the pair Sumerian /te(ga)/ ‘to attain, reach, touch upon’ with Turkic /teg/ ‘to touch, reach’ also belong in this category [Tuna,1990:24]